Abrogation of Salmonella and E. coli O157:H7 in Feedlot Cattle Fed a Proprietary Saccharomyces cerevisiae Fermentation Prototype

Salmonella and E. coli O157:H7 are insidious problems for the beef industry. Asymptomatic fecal shedding of these pathogens contaminates the hide and carcass. Furthermore, Salmonella are unique in their ability to infiltrate lymph nodes leading to the post-harvest contamination of ground beef. These contaminations yield the two most important food safety hazards associated with the consumption of beef. Herein, we report the anti-Salmonella and anti-E. coli O157:H7 effects of a novel Saccharomyces cerevisiae fermentation prototype (PRT; NaturSafeTM) fed to feedlot cattle. Cattle fed PRT were compared to those fed a combination of monensin, tylosin, and a direct-fed microbiala standard conventional practice in the U.S. beef industry. In this investigator-blinded study, 1,495 feedlot heifers (300-400 kg) were fed PRT (n=747 heifers) or the standard industry diet (PC; n=748 heifers) without PRT for 125-146 days prior to slaughter. At the abattoir, fecal swabs were obtained from 400 animals (n=200/group) and subjected to selective culture for enumerating Salmonella and E. coli O157:H7. Additionally, subiliac lymph nodes were obtained from 400 carcasses for enumeration of Salmonella spp. Salmonella isolated from the feces and lymph nodes were subjected to a virulence assay and some antibiotic susceptibility and Salmonella serovar testing. When compared to cattle that received PC, Salmonella fecal shedding, lymph node infiltration, virulence, and antibiotic resistances were significantly decreased in cattle fed PRT. Additionally, PRT-fed cattle had a lower prevalence of certain Salmonella serovars (Newport, Typhimurium, and Dublin) and shed fewer E. coli O157:H7. The decrease in Salmonella virulence was associated with a decreased expression of hilA, a genetic regulator of Salmonella invasion into eukaryotic cells. This study revealed that a proprietary Saccharomyces cerevisiae fermentation prototype inhibits the shedding, lymph node carriage, downstream virulence, and antibiotic resistance of Salmonella residing in cattle beyond the standard conventional practice that includes monensin, tylosin, and a direct-fed microbial. Abrogation of Salmonella and E. coli O157:H7 in Feedlot Cattle Fed a Proprietary Saccharomyces cerevisiae Fermentation Prototype


Introduction
Bovine-associated Salmonella are capable of causing clinical disease, but these Salmonella can also asymptomatically reside in the intestinal tract and lymph nodes of cattle. Both of these reservoirs create food safety hazards, especially the latter situation where lymph nodes serve as a protective conduit for Salmonella passage into ground beef [1]. The intestines also serve as a reservoir for E. coli O157:H7, which is nonpathogenic in cattle but highly pathogenic in humans [2].
Intestinal Salmonella and E. coli O157:H7 lead to surface contamination of beef, which can be mitigated by a number of different strategies at the abattoir [3]. Lymph node infiltration, on the other hand, is problematic given the number of lymph nodes unavoidably co-harvested with ground beef and the covertness of the Salmonella within the lymph nodes [1]. Concerns with Salmonella in the lymph nodes underscore the need for improving Salmonella prophylaxis either in the preharvest or post-harvest setting. Recent studies have demonstrated that soluble components present in Saccharomyces cerevisiae fermentation products (SCFP; SmartCare ® and Original XPC™, Diamond V) eliminate Salmonella from the gastrointestinal tract of experimentally-infected calves [4]. Specifically, SCFP (SmartCare and Original XPC) significantly reduced the intestinal burden and pathogenic effects of Salmonella in calves [4]. Because of these benefits, the anti-Salmonella effects of SCFP were examined in feedlot cattle naturally infected with Salmonella. The aims of this study were to determine a broad range of effects of a proprietary S. cerevisiae fermentation prototype (PRT; NaturSafe TM ) on Salmonella when fed to finishing cattle. Specifically fecal shedding, lymph node infiltration, downstream virulence, antibiotic resistance, and the prevalence of virulent Salmonella serotypes with a broad host range were determined. The virulence aspect was examined, in part, because SCFP (EpiCor ® , Embria Health Sciences) can increase intestinal butyrate production [5], and this short-chain fatty acid has been shown to inhibit Salmonella virulence mechanisms in vitro [6]. The antibiotic resistances were examined since SCFP (Original XPC) can modulate the microbiome [7,8], which can alter the antibiograms of resident bacteria [9]. Additionally, we examined the effects of the SCFP on the shedding of E. coli O157:H7 since this microbe shares mammalian cell adhesion strategies with Salmonella [10]. Comparisons were made to cattle fed a combination of monensin, tylosin, and a direct-fed microbial, which is a standard conventional practice in the U.S. beef industry.

Animal care and study design
Heifers (n = 1,495; 300-400kg) were obtained from two sale barns (n=438) and one backgrounding facility (n=1,057) in Kansas in May of 2015. Cattle were shipped to a commercial feedlot in Central Kansas and were provided water and hay ad libitum. On day 1 post-arrival, Journal of V Journal of Veterinary Science & eterinary Science & T Technology echnology heifers were individually weighed, identified, implanted, and vaccinated using standard procedures at the feedlot. Heifers were then randomly assigned into pens in groups of five until each pen reached its optimal capacity (~75 animals) based on bunk space and the area of the pen (14.4 inches of bunk space and 231 square feet of pen space per animal).
Two adjacent pens were designated as a single block and 10 blocks were created within the feedlot. Pens of heifers in each block were provided either a diet that contained a combination of standard industry technologies (PC), including monensin (Rumensin, Elanco Animal Health, 300 mg/head/day), tylosin (Tylovet, Huvepharma, 90 mg/head/day) and a direct-fed microbial (Bovamine Defend, Nutrition Physiology Company, 50 mg/head/day); or a diet containing 18 gm/ head/day of a S. cerevisiae fermentation prototype (PRT; NaturSafe TM , Diamond V) without monensin, tylosin, or the direct-fed microbial. Heifers received three step-up diets prior to their final finishing diet ( Table 1). All treatment feed additives were stored under manufacturerrecommend conditions and added to the final ration using a microingredient weight machine (Micro Beef Technologies, Amarillo, TX).
During the study, heifers were monitored for illness and treated as per recommendations by a veterinarian. Heifers that responded to treatment were returned to the study while non-responders were removed from the study. Morbidities and mortalities were indistinct between the two groups (data not shown).
At the conclusion of the study, heifers were shipped 145 miles to a commercial abattoir on two separate dates that were three weeks apart. These shipping dates corresponded to 125 and 146 days on study for the first and second groups, respectively. An equal number of pens per treatment group were shipped on each date (n = 5 per treatment). Shipments and carcass processing occurred on a pen-by-pen basis.

Sample collection at the abattoir
Individual animal numbers were not carried through the entire harvest process. Only the first and last animal within a pen was tagged with a lot identification tag that corresponded back to the pen number. Since individual identification was not maintained, it was not possible to keep track of which animals were sampled. As a result, lymph node and fecal samples may or may not have been collected from the same animals.
Fecal swabs were collected on the rail from 20 animals per pen (replicate). Samples were collected from every third or fourth animal within a replicate. Fecal samples were collected using a 3M-sponge stick pre-saturated with buffered peptone water. Sponge sticks were inserted into the rectum (recto-anal junction) to collect the sample. After the sample was collected, the sponge was placed into a pre-labeled bag containing buffered peptone water. The bag was closed and placed into a cooler.
Subiliac lymph nodes and the surrounding tissue were collected postevisceration. Sample collection began with the first carcass in each replicate and continued with every third or fourth carcass within that replicate. Lymph nodes were placed into pre-labeled Whirlpak bags. The bags were closed and placed in a cooler. Fecal swabs and lymph node samples were then immediately shipped on ice to Iowa State University College of Veterinary Medicine (Ames, Iowa, USA) for microbiological analyses.

Assessment of Salmonella fecal load and lymph node infiltration
Salmonella spp. were enumerated from every fecal swab sample and lymph node collected (20 per pen; 200 per treatment) using selective agar (XLD) methods described by Brewer et al. [4]. Approximately 0.3 gm of feces or lymph node were collected on a sterile cotton swab and then aseptically transferred into 10mL Lennox broth and an aliquot of the broth was immediately plated on XLD agar, incubated overnight at 37°C, and subjected to enumeration by manual counting of black-centered colonies the next day. Load was then determined as (colonies recovered) × (the dilution factor) / gm of feces or lymph node. Prevalence was calculated as percent of heifers harboring any Salmonella and was compiled across pens within a treatment group.

Assessment of E. coli O157:H7 in the Feces
E. coli O157:H7 was enumerated in 100 of the swab samples (five per pen; 50 swab samples per treatment) using selective media (Sorbitol-MacConkey agar) and a PCR targeting E. coli O157:H7 virulence genes as per Sharma and Casey [11]. Approximately 0.3 gm of feces were transferred into enrichment broth [11] and an aliquot of the broth was plated on sorbitol-MacConkey agar, incubated overnight at 37°C, and subjected to enumeration by manual counting of nonfermenting colonies the next day. From each pen-specific set of agar plates, 96 colonies were selected and subjected to the PCR targeting E. coli O157:H7 virulence genes. Load was then determined as (colonies recovered × the dilution factor × the percent of colonies yielding an E. coli O157H7-specific amplicon within the pen) / gm of feces. Prevalence was calculated as percent of heifers harboring any E. coli O157:H7 within a pen, and was compiled across pens within a treatment group.

Assessment of tissue culture invasiveness by Salmonella recovered from cattle
Approximately 50% of the recovered Salmonella were subjected to a standard antibiotic protection-based tissue culture invasion assay [12] adapted for use with both antibiotic-susceptible and antibiotic-resistant Salmonella [13]. Colonies were collected en masse, on a pen-by-pen basis, into nutrient broth and then immediately incubated for 1 hr with HEp-2 tissue culture cells at 37°C. Bacteria were recovered from inside tissue culture cells via cell lysis, incubated on XLD agar overnight at 37°C, and enumerated the next day. Percent invasion was determined as (number of black-centered colonies recovered from inside cells / number of colonies added to cells) × 100.

Assessment of virulence gene expression in Salmonella isolated from cattle
In order to correlate the virulence of Salmonella recovered from cattle with gene expression events in the pathogen, approximately  10% of the recovered Salmonella isolates were subjected to an assay that quantitates the expression of hilA (a key regulator of Salmonella invasion genes; [14]). RNA was extracted from the isolates that were collected en masse on a pen-by-pen basis, and then subjected to a semiquantitative RT-PCR targeting the hilA transcript [15].

Assessment of antibiotic resistance in Salmonella recovered from cattle
Approximately 20% of the recovered Salmonella were individually subjected to micro-broth assays with individual antibiotics (florfenicol, ceftiofur, and enrofloxacin) at breakpoint concentrations [16]. Colonies that grew in the breakpoint concentrations were deemed to be resistant. Percent resistant were then determined as (number of resistant colonies / number of colonies examined) x 100. Data were compared across pens and between groups.

Assessment of the presence of three important Salmonella serotypes in the samples
Nearly 20% of the recovered Salmonella were individually subjected to PCR assays that detect the presence of genes related to Dublin [17], Typhimurium [17], and Newport [18] serotypes. Colonies yielding a specific PCR amplicon(s) were deemed to belong to the ascribed serotype. Percent belonging to the serotype were then determined as (number of colonies yielding a specific amplicon / number of colonies examined) × 100. Data were compared across pens and between groups.

Statistical analyses
Statistical comparisons were made using an analysis of variance with Tukey's ad hoc test for multiple comparisons (GraphPad Prism, Version 6, La Jolla, CA). Significant differences were defined at P ≤ 0.05.

Assessment of Salmonella Fecal Shedding in Cattle
Salmonella can be a potential pathogen source for humans who consume improperly cooked beef. Carcasses can become exposed to Salmonella through fecal contamination during processing. In this study, fecal shedding of Salmonella was evaluated in 200 heifer's postmortem from each treatment group. As shown in Figure 1a, fecal shedding of Salmonella was significantly less (P<0.05) in cattle fed PRT (105 versus 405 CFU/gm of feces, respectively). The relative prevalence of fecal shedding was significantly less (P<0.05) in heifers fed PRT (6 versus 13%, respectively) as per Figure 1b.

Assessment of lymph node infiltration by Salmonella
Since Salmonella lymph node carriage can be a source of contamination of ground beef [19], Salmonella load was determined in the subiliac lymph nodes of 200 carcasses from each treatment group. As shown in Figure 2a, lymph node infiltration was significantly less (P<0.05) in carcasses from heifers fed PRT (902 versus 6,642 CFU/gm of lymph node, respectively). The percent of Salmonella-bearing lymph nodes was significantly less (P<0.05) in carcasses from heifers fed PRT (4 versus 14%, respectively) as per Figure 2b.

Determination of E. coli O157:H7 in the feces of cattle
To determine if PRT had an effect on the presence of E. coli O157:H7 in the feces of the heifers, fecal samples (100 per treatment group) were quantitatively examined for the presence of this critical foodborne pathogen. As shown in Figure 3a, heifers fed PRT had a statistically lower (P<0.05) E. coli O157:H7 fecal load than heifers fed PC (52 versus 122 CFU/gm of feces, respectively). Figure 3b reveals a decreased prevalence (P<0.05) of E. coli O157:H7 in heifers fed PRT when compared to those receiving the PC diet (37 versus 57%, respectively).

Assessment of the virulence of Salmonella recovered from cattle
In order to compare the virulence of Salmonella recovered from cattle, the isolates were subjected to an assay that predicts the ability of Salmonella to invade gut epithelial cells, which is a hallmark of Salmonella virulence [20]. The effects of PRT on virulence were examined due to the ability of SCFP to increase butyrate in the intestine [5]. Research has shown that butyrate can decrease the Salmonella invasion gene (hilA) expression in vitro [6], which results in the decreased ability of Salmonella to invade cells. In the current study, invasiveness of Salmonella was significantly less (P<0.05) in Salmonella recovered from the feces and lymph nodes of cattle fed PRT (Figure 4). This decrease in invasiveness coincided with a decrease in the expression of hilA ( Figure  5), a major regulator of Salmonella virulence for mammalian hosts [14].

Assessment of the antibiotic resistance of Salmonella recovered from cattle
The antibiotic resistance of Salmonella in lymph nodes and fecal samples were examined since SCFP can modulate the microbiome [7,8], which can alter the antibiograms of resident bacteria [9]. To assess the possibility that PRT inhibits antibiotic resistant Salmonella or induces the expulsion of antibiotic resistance elements from Salmonella, isolates      recovered from cattle were subjected to an antibiotic susceptibility assay that utilized three individual antibiotics (ceftiofur, enrofloxacin, and florfenicol). These three antibiotics were chosen given their extended spectra and importance in bovine therapeutics. Additionally, two of the three antibiotics tested (ceftiofur and enrofloxacin) have counterparts important for human therapeutics (ceftriaxone and ciprofloxacin, respectively). Figure 6 reveals a decrease in the prevalence of resistant (P<0.05) Salmonella recovered from heifers fed PRT for all three antibiotics. This figure represents isolates from both feces and lymph nodes. It is of note that resistance of these antibiotics was, in general, more prevalent in the fecal isolates when compared to the lymph node isolates.

Assessment of the presence of three important Salmonella serotypes in the samples
Salmonella isolates recovered from the feces or lymph nodes of cattle were subjected to PCR assays targeting three serotypes (Dublin, Newport, and Typhimurium). As shown in Figure 7, the prevalence of two of these serovars was diminished in heifers fed PRT, regardless of the source of the isolates. No S. Dublin were isolated from feces and only one colony of S. Dublin was isolated from lymph nodes. Thus, statistical evaluations are not presented for this minor subsection of the study.

Discussion
Salmonella and E. coli O157:H7 are insidious problems for the beef industry and represent critical food safety hazards. Salmonella and E. coli O157:H7 can be shed in fecal material that can contaminate the carcass during processing. Salmonella is also harbored in the lymph nodes, which can lead to contamination of ground beef. Therefore, identifying mitigation strategies for both pathogens is needed especially considering the covert nature of Salmonella lymph node infiltration [19].
In this study, the anti-Salmonella and anti-E. coli O157:H7 effects of NaturSafe TM (PRT) were examined and two critical indicators of Salmonella contamination (fecal shedding and lymph node infiltration) were significantly reduced by NaturSafe TM . In this study, heifers fed PC shed a higher number of Salmonella and E. coli O157:H7 and had more Salmonella present in the lymph nodes, which ultimately increases the risk of pathogen transmission to humans that ingest beef. There are currently no other research studies looking at the effects of NaturSafe TM on pathogens in beef cattle but there are studies examining the effects of other SCFP products on pathogen load. These results described herein are consistent with a previous study in which SCFP (SmartCare and Original XPC) reduced intestinal colonization in pre-ruminant calves [4], which is possibly due to macrophage-associated clearance of the pathogen. In another study, cannulated beef heifers supplemented with SCFP (Original XP ® , Diamond V) had less E. coli O157:H7 colonizing the recto-anal junction after being challenged with E. coli O157:H7 [21]. Therefore, SCFP products seem to minimize the levels of Salmonella and E. coli O157:H7 in cattle.
Other significant and unique findings in this study were the reduction in virulence and antibiotic resistance in Salmonella recovered from heifers fed PRT. Reduced virulence was detected by diminished tissue culture invasion with a concomitant reduction in the expression of hilA [14]. It is unclear how virulence was mitigated by feeding this SCFP prototype. However, previous studies revealed that SCFP (EpiCor) increases the gastrointestinal production of butyrate [5], a putative repressor of hilA expression [22]. Regardless of the mechanism underlying this effect, the observed magnitude of decreased invasiveness is likely to increase the infectious dose of Salmonella for a human as evidenced by our prior study, in which this level of diminished invasiveness altered the murine LD 50 approximately 5-fold [13].

Conclusions
In summary, NaturSafe TM -fed feedlot heifers were significantly less likely to shed Salmonella and harbor this pathogen in the lymph nodes. The anti-shedding effect of NaturSafe TM was also observed for E. coli O157:H7. Additionally, NaturSafe TM reduced the virulence and antibiotic resistance of recovered Salmonella. Ultimately, these beneficial effects will have a marked positive effect on food safety in the beef industry. Future mechanistic studies are warranted to uncover the molecular bases for these effects.

����������
This work was funded by Diamond V Salary support was provided to K.L.A. in exchange for her time devoted to the project, but no other financial compensation was or will be awarded for this study. Thus the authors declare no conflict of interest.